Lait (1997) 77, 561-573© Elsevier/lnra

561

Original article

Effect of compositional and environmental factorson the growth of indigenous non-starter lactic acid

bacteria in Cheddar cheese*

eN Lane**, PF Fox, EM Walsh, B Folkertsma, PLH McSweeney

Department of Food Chemistry, University College, Cork. Ireland

Summary - The effects of salt and moi sture content, ripening temperature and starter strain, inde-pendently, on the growth of indigenous non-starter lactic acid bacteria (NSLAB) in Cheddar cheeseswere studied. Salt-in-moisture (SIM) :,; 5.9% had no effect on the growth of NSLAB; 6.1 % SIMreduced their growth rate but had no effect on their number at 6 months. Cheeses with a moi sture con-tent > 40.3% had higher numbers of NSLAB than cheeses with a lower moisture content during theinitial stages of ripening; the difference appeared to be related to the length of the lag phase rather thanto the rate of growth of these bacteria. Numbers of NSLAB were quite low in cheeses ripened at4 "C even after 6 months and increased with ripening temperature at ail stages of ripening. NSLABgrew faster initially in cheeses made with Lactococcus lactis ssp lactis ML3 or 303 as starter than inthose made with Le lactis ssp cremoris AM 1 or AM2 but reached similar numbers after 6 months.

non-starter lactic acid bacteria 1 Cheddar cheese 1 Cheddar cheese composition

Résumé - Effet des facteurs composition et environnement sur la croissance des bactérieslactiques indigènes non levains dans le fromage cheddar. Les effets respectifs de la teneur en selet en eau, de la température d'affinage et des souches de levains lactiques sur la croissance des bac-téries lactiques indigènes non levains (NSLAB) ont été étudiés dans du fromage de cheddar. Le rap-port sel/humidité (StH), inférieur ou égal à 5,9 %, n'a pas d'effet sur la croissance des NSLAB, uneteneur à 6,1 % de StH réduit leur taux de croissance mais n'a pas d'effet sur leur nombre à 6 mois d'affi-nage. Les fromages ayant un taux d'humidité supérieur à 40,3 % ont un plus grand nombre deNSLAB que les fromages ayant un taux d'humidité plus faible durant la phase initiale d'affinage. Ladifférence semble être liée à la durée de la phase de latence, plutôt qu'au taux de croissance de ces bac-téries. Les nombres de NSLAB sont plutôt bas dans des fromages affinés à 4 "C, même après 6 moisd'affinage, et croissent en même temps que la température à tous les stades de l'affinage. Le taux deNSLAB augmente initialement plus vite dans des fromages faits avec des Lactococcus lactis spplactis ML3 ou 303 utilisés en tant que levains comparés à ceux fabriqués avec des Lactococcus lac-tis spp cremoris AMI ou AM2, mais atteignent le même taux après 6 mois d'affinage.

bactérie lactique non levain 1 fromage 1 cheddar

* Oral communication at the !DF Symposium 'Ripening and Quality of Cheeses', Besançon,France, February 26-28, 1996** Correspondence and reprints

562 CN Lane et al

INTRODUCTION

Non-starter lactic acid bacteria (NSLAB)in Cheddar cheese are adventitious micro-organisms, introduced mainly by post-pas-teurisation contamination and capable ofgrowing in chee se during ripening. In con-trast to starter bacteria, the numbers of whichdecrease after manufacture at a rate which isstrain-dependent, NSLAB grow from lownumbers « 101 cfu/g) in fresh curd to dom-inate the viable microflora of mature Ched-dar cheese. Mesophilic lactobacilli are thepredominant NSLAB in Iris'h Cheddar(Jordan and Cogan, 1993); strains typicallyfound include Lb casei, Lb plantarum, Lbpseudoplantarum, Lb curvatus and Lb bre-vis. Since lactobacilli have a complex pro-teolytic system (Peterson and Marshall,1990; Kanawjia et al, 1993), they may playa role in proteolysis during cheese matura-tion and hence affect the quality of the finalproduct. While many authors agree thatNSLAB may affect the flavour of cheese,sorne consider this effect to be undesirable(Kleter, 1977) while others suggest that thesebacteria contribute to the development of astrong flavour which may be attractive tosorne consumers (McSweeney et al, 1993).NSLAB have also been associated withdefects such as calcium D-Iactate crystalson the cheese surface (Khalid and Marth,1990).

Since NSLAB grow in cheese duringripening, while the numbers of most otherbacteria decline, cheese must contain sub-strates for their growth. Potential growthsubstrates include lactose, lactate, citrate,sugars from enzymatic hydrolysis of J(-

casein, pyruvate, free amino acids, peptidesand starter cell lysate (Peterson andMarshall, 1990; Martley and Crow, 1993).However, the actual substrates used byNSLAB in Cheddar cheese have not beenelucidated. In addition to available sub-strates, it is probable that cheese composition(ie, pH, salt concentration, moisture con-tent) influences the rate and extent of growth

of NSLAB; however, very little research onthis subject has been reported. Temperatureis a major factor controIling bacterial growthin ail cheese types. The increased rate ofgrowth of NSLAB in Cheddar cheese atelevated ripening temperatures has beendemonstrated (Cromie et al, 1987;Folkertsma et al, 1996).

The effects of the salt and moi sture con-tent of Cheddar cheese, ripening temperatureand starter strain on the growth of indige-nous NSLAB during ripening were inves-tigated in this study.

MATE RIALS AND METHODS

Cultures

Lactococcus lactis ssp cremoris 223, Le lactisssp cremoris AM 1, Le lactis ssp cremoris AM2,Le lactis ssp lactis ML3 (a1so named NCDO 763)and Le lactis ssp lactis 303 were obtained fromthe culture collection of the Department of Micro-biology, University College, Cork. Frozen cul-tures (1%, v/v) were propagated at 30°C for 16 hin autoclaved (110°C, 10 min) reconstitutedskim-milk powder (100 g L-1). Two transferswere made prior to cheesemaking; bulk cultureswere grown for 16 h at 21°C.

Standard cheesemaking protocol

Cheddar cheese was made in a pilot-plant byconventional methods from pasteurised (74 "C,15 s) bulk milk using 2% (v/v) Le lactis ssp cre-moris 223 as starter (unless otherwise stated).Calf rennet (Chr Hansen's Laboratory, LittleIsland, Cork) was added to the cheesemilk at alevel of 0.3 mL L-1. The setting temperature was30 "C and the coagulum was eut - 50 min afterrenneting. The curds were cooked at 39 "C andcheddared until the pH reached 5.3-5.4. Unlessotherwise stated, milled curds were salted at alevel of2.5% (w/w), pressed ovemight at 1 bar at16 ± 1°C, vacuum packed and ripened at 8 "C for6 months. Ali cheese trials were carried out induplicate.

The cheesemaking protocol was modified tomanufacture cheeses with varying salt or mois-

Growth of indigenous NSLAB in Cheddar chee se

ture content, with different starter strains or dif-ferent ripening temperature as follows.

Salt content

Curd ( 10 kg) prepared as above was divided aftercheddaring into six 1.7-kg lots and salted at alevel of 1.5, 2.1,2.7,3.3,3.9 or 4.5% (w/w).

Moisture content

It was found that the most successful method forpreparing cheese of different moisture contentwas to salt the curd at different pH values; 2-kglots of curd were taken at pH 5.9, 5.7, 5.5, 5.3or 5.1, milled, salted at 2.5% (w/w) and pressedas described above.

Ripening temperature

Curd was divided into three 2-kg blacks, salted at2.5% (w/w), pressed overnight at 16 ± 1°C, vac-uum packed and ripened at 4, 8 or 12 oc.

Starter strain

Chee se (2 kg) was made by the conventionalmethod described above using 2% (v/v) Le lac-tis ssp cremoris AMI, Le lactis ssp cremoris

563

AM2, Le lactis ssp lactis ML3 or Le lactis ssplactis 303 as starter.

Bacteriological analysis

Total bacteria in the cheeses were enumeratedon LMI7 agar (Terzaghi and Sandine, 1975)after incubation at 30 "C for 3 days. Growth ofNSLAB in the cheeses during ripening was esti-mated by enumeration on Rogosa agar (Rogosaet al, 1951) after incubation at 37 "C for 5 days.

Compositional analysis

Samples of each cheese were taken immediatelyafter pressing for duplicate determination of pro-tein (IDF, 1964), fat (URS, 1955), moi sture (IDF,1982) and salt (Fox, 1963). The pH of gratedcheese (lOg) macerated in 10 mL distilled waterwas measured.

RESULTSEffect of salt

The composition of the experimental Ched-dar cheeses is shown in table I.During

Table I.Composition] of l-day-old experimental cheeses manufactured to give varying salt contents.Composition 1 de fromages expérimentaux à l jour, fabriqués avec différentes teneurs en sel.

Salting level Salt Moisture SIM2 Fat Protein pH(%) (%) (%) (%) (%) (%) (%)

Trial}1.5 1.13 39.6 2.85 29.8 25.0 5.232.1 1.38 38.6 3.58 30.0 24.7 5.262.7 1.59 39.5 4.03 29.5 24.7 5.323.3 1.70 39.9 4.26 30.0 25.0 5.333.9 1.77 38.5 4.60 29.8 25.3 5.344.5 2.05 38.2 5.37 29.8 25.6 5.37

Trial 21.5 1.19 38.6 3.08 33.0 24.7 5.182.1 1.57 37.3 4.21 33.0 25.3 5.272.7 1.84 35.7 5.15 33.0 26.1 5.333.3 1.94 36.1 5.37 34.3 25.7 5.343.9 2.04 34.4 5.93 33.5 26.5 5.394.5 2.10 34.5 6.09 33.0 26.6 5.39

] Mean of duplicates. 2 Salt-in-moisture,

J Moyenne de deux déterminations. 2 Rapport sellhumidité.

-------------------------~~ ~------~~~~~-

564 CN Lane et al

manufacture, curds were salted at levelsfrom 1.5 to 4.5% (w/w) at 0.6% increments,resulting in cheeses with a salt content rang-ing from 1.13 to 2.05% (trial 1) and from1.19 to 2.10% (trial 2). The moisture contentof the chee ses generally decreased withincreasing salt content (especially in trial2). Since the concentration of salt in theaqueous phase, rather than the total salt con-tent, is an important parameter for thegrowth ofbacteria in cheese, the salt contentwas expressed as salt-in-cheese moisture

(SIM). The pH of the cheeses increased withincreasing salt content, ranging from 5.23to 5.37 in trial 1 and 5.18 to 5.39 in trial 2.

Starter bacteria were enumerated onLM 17 agar but NSLAB can also grow onthis medium and, if present in sufficientlylarge numbers, will cause overestimation ofstarter bacteria. To avoid this, counts onLM 17 during the early stages of ripeningonly (when the numbers of NSLAB werelow) were considered to be a val id estima-tion of starter numbers. Immediately after

Table II. Changes in starter population in experimental cheeses during the early stages of ripening.Changement de population des levains dans des fromages expérimentaux pendant les premièresphases d'affinage.

Cheese type Ripening time (weeks)0 1 3 4 0 3 4

Trial 1 Trial 2log cfu r' cheese

Salting level (%)1.5 8.0 7.3 7.6 7.2 8.0 7.2 6.4 6.02.1 8.1 6.8 7.3 7.2 8.0 6.9 5.9 5.82.7 8.2 7.0 7.0 7.0 8.0 6.9 6.1 5.83.3 8.2 7.0 7.0 6.4 8.1 7.3 6.0 6.03.9 8.0 7.0 6.3 6.1 8.0 5.5 5.6 6.04.5 8.0 7.4 6.6 6.1 8.0 6.2 5.6 6.4

Milling pH5.1 7.2 ND 6.5 6.3 7.2 6.9 6.5 6.45.3 7.5 ND 6.9 6.4 7.9 6.5 6.3 6.65.5 7.7 ND 6.7 6.4 7.7 6.9 6.5 6.85.7 7.2 ND 6.7 6.2 7.3 6.8 7.1 8.05.9 7.4 ND 6.8 6.3 7.9 7.0 7.9 8.3

Ripeningtempe rature (OC)

4 8.0 7.0 6.5 5.9 7.1 6.9 6.4 5.38 8.0 6.8 6.3 5.9 7.1 7.0 7.3 6.512 8.0 6.5 6.3 5.4 7.1 6.5 5.9 5.6

Starter strainAMI 5.6 5.6 6.5 6.0 6.4 6.7 6.0 6.2AM2 5.4 ND 4.8 4.5 4.7 4.6 ND 5.7ML3 9.1 9.2 7.6 8.1 8.6 7.2 ND 7.4303 9.6 9.9 9.6 9.5 9.6 9.6 9.6 9.4

ND, not determined.

ND: non déterminé.

Growth of indigenous NSLAB in Cheddar cheese 565

pressing, starter counts were between 1.0and 1.5 x 108 cfu s' cheese in trial 1 and1.0 and 1.2 x 108 cfu g-I in trial 2 and therate at which they decreased over the nextfew weeks appeared to be independent ofthe SIM in the cheeses (table II).

NSLAB were not detected in the cheesesat one day but grew in ail cheeses duringripening, reaching numbers from 4.8 x 107

to 2.5 x 108 cfu g-I in trial 1 and from3.5 x 107 to 1.5 X 108 cfu g-I in trial 2 after6 months (fig lA, B). SIM levels of2.85 to5.37% in the experimental chee ses in trial 1had no effect on the rate at which NSLABgrew in the cheese (fig 1A). In trial 2, dif-ferences in NSLAB numbers during ripen-ing in cheeses with SIM levels ::;5.93% didnot appear to be related to increasing SIM

109A ---0- 2.8S%S/M

108 --0-- 358%5/M--l:r- 4.03%5/M

107 - 426%5/M

Qj' ---+- 4.6Q%5/M<Il 106 ----.- 5.37%5/MQIQI

..<::~ 105--.E

104~l'Q< 103...l

~102

101

10°0 4 8 12 16 20 24

Time (wks)

109B 3.08%5/M

108 4.21%5/M5.15%5/M

107 5.37%5/M

Qj' ---+- 5.93%5/M<Il 106 ----.- 6.Q9%5/MQIQI

-5105se--.E104~

l'Q-e103...l

<J'lZ

102 Fig 1. Effect of salt-in-moisture (SIM) con-

101tent on the growth of non-starter lactic acidbacteria during ripening. A. Trial 1. B. Trial

10°2. Note: 10° "'< 10 cfu NSLAB g-I cheese.

0 4 8 12 16 20 24 Effet de la teneur sel/humidité sur la crois-sance des bactéries lactiques non levains

Time (wks) pendant l'affinage. A. Essai J. B. Essai 2.

566 CN Lane et al

content (Fig lB); in the cheese with thehighest SIM level (ie, 6.09%), NSLAB grewslowly to 5.5 x 104 cfu g-l at 4 months.NSLAB in all cheeses in trial 2 were similarat 6 months (fig lB).

Effeet of moisture

The composition of the experimental Ched-dar cheeses is shown in table III. The rangeof moisture levels was higher in trial 2(39.2-45.0%) than in trial 1 (38.4-43.7%).The pH of the cheeses increased withincreasing moisture content, probably dueto the fact that, during manufacture, thecurds were milled at different pH values sothat cheeses varying in moisture contentcou Id be produced. Salt content variedbetween the cheeses in each trial. The levelsof fat and protein in the cheeses decreasedwith increasing moi sture content (table III).

Starter numbers in the cheeses expressranges from 1.7 to 4.5 X 107 cfu g-I in trial1 and from 1.4 to 7.7 x 107 cfu s' in trial 2

and generally decreased by approximatelyone log cycle after 4 weeks (table II). Thehigh starter numbers after 4 weeks ripeningin the cheeses in trial 2 with a moi sture con-tent of 42.9 or 45.0% mayindicate contami-nation by non-starter micro-organisms;NSLAB counts in the se cheeses were1.5 x 107 and 1.9 x 106 cfu g', respectively,after 4 weeks (fig 2B).

The effect of moisture content on thegrowth of NSLAB in the cheeses duringripening is shown in figure 2A, B. Thegrowth of NSLAB in cheeses with a highmoisture content (39.7-43.7% in trial l,41.3-45.0% in trial 2) had a shorter lagphase than in chee ses with lower moisturecontent but after the lag period, the rate atwhich NSLAB grew appeared to be inde-pendent of moi sture content. NSLAB num-bers in all cheeses in trial 1 reached similarvalues after 3 months and were maintainedthroughout the latter half of ripening(fig 2A). Although NSLAB numbers var-ied between the five cheeses in trial 2between 2 and 4 months, this did not appear

Table III. Composition 1 of I-day-old experimental cheeses manufactured to gi ve varying moi sturecontents.Composition' defromages expérimentaux à J jour,fabriqués avec différentes teneurs en eau.

Milling pH Salt Moisture SIM2 Fat Protein pH(%) (%) (%) (%) (%)

Trial J5.1 1.36 38.4 3.54 30.5 26.1 5.235.3 1.63 38.5 4.23 30.0 25.5 5.305.5 1.73 39.7 4.36 29.5 25.3 5.385.7 1.66 41.4 4.01 28.0 24.5 5.375.9 1.69 43.7 3.87 26.8 23.9 5.44

Trial 25.1 1.47 39.2 3.75 33.3 26.0 5.105.3 1.48 40.3 3.67 29.5 25.2 5.125.5 1.38 41.3 3.34 29.0 24.6 5.155.7 1.37 42.9 3.19 29.0 24.2 5.175.9 1.68 45.0 3.73 26.8 23.2 5.23

1 Mean of duplicates. 2 Salt-in-moisture.1 Moyenne de deux déterminations. 2 Rapport sel/humidité.

Growth of indigenous NSLAB in Cheddar cheese 567

109 A ---0- 38.4%M

108 -0- 385%M--- 39.7%M

107 ~ 41.4%M

Q;' -.- 43.7%M<Il 106cucu

,.<:;v

105oc--..2104~

~-c 103....JCf)

Z102

101

10°0 4 8 12 16 20 24

Time (wks)

109

B ---0- 39.2%M

108 -0- 40.3% M--- 41.3%M

107 ~ 42.9%M

Q;' -.- 4S.Q%M<Il 106cucu

,.<:;v 105oc--..2 104~~<t:

103....JCf)

Z102

Fig 2. Effect of moisture content on the

101 growth of non-starter lactic acid bacteriaduring ripening. A. Trial 1. B. Trial 2. Note:

10° 10° "'< 10 cfu NSLAB s' cheese.0 4 8 12 16 20 24 Effet de la teneur en eau sur la croissance

des bactéries lactiques non levains pendantTime (wks) l'affinage. A. Essai 1. B. Essai 2.

Effeet of ripening temperatureto be related to moisture content (fig 2B).NSLAB reached similar numbers in aIlchee ses in trial 2 after 6 months ripening(fig 28).

The composition of the experimental Ched-dar cheeses is shown in table IV. The salt

568 CN Lane et al

Table IV. Composition 1 of I-day-old experimental cheeses ripened at different temperatures.Composition' de fromages expérimentaux à 1jour, affinés à différentes températures.

Ripening Salt Moisture SIM2 Fat Protein pHtemp (oC) (%) (%) (%) (%) (%)

Trial l4 1.43 39.6 3.61 31.0 24.5 5.238 \.50 39.4 3.81 31.0 24.5 5.2612 1.47 40.4 3.64 31.0 23.9 5.26

Trial 24 1.27 38.4 3.31 30.0 25.1 5.238 1.48 37.6 3.94 29.5 25.0 5.3012 1.36 38.3 3.55 30.0 24.4 5.26

1 Mean of duplicates. 2 Salt-in-moisture.

J Moyenne de deux déterminations. 2 Rapport sel/humidité.

content of the cheeses was quite low con-sidering that the curds were salted at a levelof 2.5% (w/w). The moisture content of thecheeses in trial 1 was slightly higher thanwould be expected for commercial Ched-dar cheese.

Starter counts in cheeses ex-press wereidentical within trials 1or 2 at I.I x lOScfu s:'or lA x 107 cfu s'. respectively, but theydecreased at different rates (table II). In triall, the cheese ripened at 12 "C had the low-est number of starter at ail sampling timesand had decreased to 2.3 x 105 cfu g-l after4 weeks; the rate of decrease of starter num-bers was faster at 8 "C than at 4 "C duringthe very early stages of ripening (up to 3weeks) but these cheeses had similar num-bers of viable starter cells (7.8 and7.3 x lOScfu s'. respectively) after 4 weeks(table II). The trend observed in cheeses intrial 1 was not repeated in trial 2 in which thecheese ripened at 8 "C had higher starternumbers than those ripened at 4 or 12 "C atail sampling limes, and decreased to3.3 x 106,2 x 105 and 4 x 105 cfu g",respectively after 4 weeks (table II).

NSLAB generally grew in the se cheesesin the order 12 > 8 > 4 "C (ie, the higher theripening temperature, the higher the num-

bers of NSLAB in the cheese) and reached2.0 x 108,2.7 X 105 and 4.0 x 104 cfu g-l,respectively, in trial 1 and 2 x 107, 1.8 X 107

and 1.3 x 105 cfu g', respectively, in trial 2,after 6 months (fig 3A, B). The effect oftemperature on the growth rate of NSLAB incheese in trial 1 ripened at 8 "C was notasgreat in the corresponding cheese in trial 2(fig 3A, B).

Effeet of starter strain

The composition of the experimentalcheeses is shown in table V. The pH of thecheeses made with strains of Le lactis ssplactis ML3 or 303 was lower than thosemade with strains of Le lactis ssp cremorisAM 1 or AM2, with the exception of thecheese in trial 2 made with Le lactis ssp lac-tis ML3. The moisture content of the cheeseswas higher than expected for commercialCheddar cheese but the salt, fat and proteincontents of the cheeses were generallywithin nonnallimits.

The cheeses in both trials made witheither Le lactis ssp cremoris AM 1 or AM2had much lower starter numbers than thosemade with Le lactis ssp lactis MU or 303 at

Growth of indigenous NSLAB in Cheddar cheese 569

109A

108

107

QI106<J)

'"'"..<::~ 105---= 104~

co:) 103enZ

102

101

1000 4 8 12 16 20 24

Time (wks)

109B

108

107

QI 106<J)

'"'"..<::

~ 105---= 104~co<t: 103....l Fig 3. Effect of ripening temperature COC)enZ

102 on the growth of non-starter lactic acidbacteria during ripening. A. Trial 1. B.

101 Trial 2. Note: 10° "'< 10 cfu NSLAB g-Icheese.

100 Effet de la température d'affinage (OC)

0 4 8 12 16 20 24 sur la croissance des bactéries lactiquesnon levains pendant l'affinage. A. Essai J.

Time (wks) B. Essai 2.

one day (table II). The change in numbers ofAM 1 and AM2 during the first 4 weeks ofripening varied unpredictably, possibly dueto the growth of NSLAB on the LM 17 agar,especial1y since the number of starter in

the se cheeses was quite low; NSLAB num-bers reached 3.6 x 104 cfu g-I in trial 1 and1.1 x 103 cfu s' in trial 2 chee ses madewith strain AM 1 or 1.2 x 103 cfu g-I in trial1 and 5.4 x 104 cfu s' in trial 2 cheeses

570 CN Laneet al

Table V. Composition 1 of l-day-old experimental chee ses made with different starter strains.Composition 1 de fromages expérimentaux à l jour, fabriqués avec différentes souches de levain.

Starter strain Salt Moisture SIM2 Fat Protein pH(%) (%) (%) (%) (%)

Trial}AMI 1.33 40.7 3.27 32.3 24.4 5.25AM2 1.41 40.5 3.48 31.5 24.0 5.\7MU 1.39 43.4 3.20 30.5 24.6 5.02303 1.23 44.1 2.79 30.3 23.2 5.00

Trial 2AMI 1.53 42.1 3.63 30.3 24.0 5.23AM2 1.62 41.1 3.94 32.0 24.3 5.29MU 1.79 42.0 4.26 31.0 22.7 5.35303 1.50 41.9 3.58 30.8 22.8 5.04

1 Mean of duplicates. 2 Salt-in-moisture.1 Moyenne de deux déterminations. 2 Rapport sel/humidité.

made with strain AM2 over this ripeningperiod. The number of ML3 in cheese inboth trials decreased by approximately onelog cycle to 1.2 x 108 and 2.7 x 107 cfu g-I,respectively, while the numbers of strain303 in both trials remained high and were3.5 x 109 and 2.4 x 109 cfu g-l, respectively,after 4 weeks (table II).

The growth of NSLAB in the cheesesduring ripening is shown in figure 4A, B.In general, the rate of NSLAB growth incheese made with either Le laetis ssp lac-tis MU or 303 was higher during the earlystages of ripening than in chee se made withLe laetis ssp cremoris AM 1 or AM2.NSLAB numbers in ail four cheeses in trial1 reached similar values after 3 months andincreased only slightly between 3 and 6months (fig 4A). The same trend was foundin trial 2 for the cheeses made with strainsAM 1, AM2 or MU but NSLAB numbers incheese made with strain 303 were lowerthan in the other cheeses at both 3 and 4months (fig 4B). NSLAB numbers in ailcheeses in trial 2 were similar at 6 months.

DISCUSSION

The effect of sorne compositional and envi-ronmental parameters on the growth ofindigenous non-starter lactic acid bacteria(NSLAB) in Cheddar cheese made frompasteurised milk was studied. High salt-in-moisture (SIM) levels appeared to have aninhibitory effect on the activity of the starterbacteria, as indicated by a slightly high pH.An inverse relationship between pH and saltcontent of chee se has been reported(O'Connor, 1974; Turner and Thomas,1980; Thomas and Pearce, 1982). There wasa decrease in viable starter numbers in theexperimental chee ses during the first fewweeks of ripening. These results are inagreement with the findings of many others,including Visser (1977) and McSweeney etal (1993). The viability of starter cells dur-ing the initial stages of ripening appearedto be independent of SIM in the chee se. SIMvalues of 2.85-5.93% (trials 1 and 2) didnot effect the growth of indigenous NSLABduring ripening; although 6.09% SIM

Growth of indigenous NSLAB in Cheddar cheese 571

109A --0-- AM!

108 --0-- AM2-.- ML3

107 -e-- 303..106III.,.,

..<:u 1050.0--..2 104~~<t: 103...:J

~102

101

10°0 4 8 12 16 20 24

Time (wk)

109B --0-- AM!

108 --0-- AM2-.- ML3

107 -e-- 303..III 106.,.,

..<:

~ 105--..2 104~~<t: 103...:J(fl

Z Fig 4. Effect of starter strain on the102growth of non-starter lactic acid bacte-

101 ria during ripening. A. Trial 1. B. Trial2. Note: 10° =< 10 cfu NSLAB g-I

10° cheese.0 4 8 12 16 20 24 Effet des souches de levain sur la crois-

sance des bactéries lactiques non levainsTime (wk) pendant l'affinage. A. Essai J. B. Essai 2.

reduced the growth of NSLAB up to 4months, the final numbers were not affectedby the SIM. Jordan and Cogan (1993), whoexamined the salt tolerance of NSLAB (Lbcasei, Lb plantarum and Lb curvatus) iso-

lated from commercial Irish Cheddar cheese,found that ail isolates grew in the presenceof 4% NaCI, most grew in 6% and sornegrew at salt concentrations as high as 10%.They concluded that the levels of SIM nor-

572 eN Lane et al

mally found in Cheddar cheese (ie, 4-6%)wou Id have little inhibitory effect on thegrowth of NSLAB.

Within the range studied, moi sture con-tent did not influence the decline of starterbacteria during the early stages of ripeningand did not appear to influence the rate orfinal numbers of NSLAB although theirgrowth in chee ses with high moi sture levelshad a shorter lag phase than those with alower moisture content. It is not possible toattribute any differences found betweenNSLAB counts to the level of moisturealone since the method used to make cheesevarying in moi sture content caused sub-stantial variations in other compositionalfactors, especially pH (and hence the amountof unfermented lactose), and these mayaffect the growth of indigenous NSLAB.

The effect of ripening temperature on thedecline of starter numbers during the initialstages of ripening is unclear since the trendfound in trial 1 was not repeated in trial 2.Law et al (1979) found that starter numbersdecreased more slowly in cheese ripened at8 than at 13°C while Cromie et al (1987)reported that starter counts were lower inchee se stored at 8 "C than at 15, 17.5 or20 oc. NSLAB numbers in cheeses ripenedat 4 "C were quite low even after 6 monthsand increasing the ripening temperatureresulted in increased numbers of NSLAB.Similar results were found by Cromie et al(1987) for cheeses ripened at 8, 15, 17.5 or20 oc. Folkertsma et al (1996) found thatripening temperature and, in particular, therate at which the cheese was cooled afterpressing had considerable effects on thegrowth of NSLAB during ripening althoughripening temperature (8, 12 or 16°C) hadlittle effect on the number of NSLAB after9 months.

Numbers of starter bacteria in cheesesmade using Le lactis ssp cremoris AM 1 orAM2 were low at 1 day, possibly due to thefact that the se strains are quite sensitive tocooking temperature and salt (Martley and

Lawrence, 1972; Fedrick et al, 1986). Highstarter numbers were observed at 1 day incheeses made with Le lactis ssp lactis ML3or 303 and the rate of dec!ine of these bac-teria was very slow during the early stagesof ripening. These results are in agreementwith the findings of others, includingChapot-Chartier et al (1994) and Morgan etal (1995). The rate at which starter cells dieand lyse may influence the growth ofNSLAB (Thomas, 1987), either directly byproviding growth substrates or indirectly byreleasing intraceIIular peptidases which Iibe-rate amino acids as growth substrates forNSLAB. Furthermore, the rapid decline ofstrains AM 1 and AM2 may have led tounfermented lactose which is a potentialsubstrate for NSLAB growth. However, thepresent results indicate that indigenousNSLAB grew more rapidly during the initialstages of ripening in cheese made with Lelactis ssp lactis ML3 or 303 than in cheesemade with Le lactis ssp cremoris AM 1 orAM2 but the starter strain had !ittle effecton the final numbers of NSLAB in thecheeses. Hence, further research on the influ-ence of starter strain (and starter Iysis) onthe growth of NSLAB appears warranted.

The fact that NSLAB grow during ripen-ing and may affect, positively or negatively,the overall quality of cheese has encouragedinvestigation into the factors that affect thegrowth of these micro-organisms and thesubstrates available to them in cheese. Theresults of this study indicate that normalvariation in the salt-in-moisture (4-6%) andmoisture content « 39%) of Cheddar cheesewould have little effect on the growth ofindigenous NSLAB. The temperature atwhich cheese is ripened influences the rateof growth of NSLAB and this should betaken into account when suggesting tem-perature as a means of accelerating theripening process. The initial growth ofindigenous NSLAB appears to be starter-dependent but further research is requiredon this subject.

Growth of indigenous NSLAB in Cheddar chee se

ACKNOWLEDGMENTS

This project was funded in part From structuralfunds From the Food Sub-programme of the EUIndustry programme.

REFERENCES

Chapot-Chartier MP, Deniel C, Rousseau M, VassalL, Gripon JC (1994) Autolysis of two strains ofLaetocoeeus laetis during cheese ripening. Int DairyJ 4,251-269

Cromie SJ, Giles JE, Dulley JR (1987) Effect of ele-vated ripening temperature on the microllora ofCheddar cheese. J Dairy Res 54, 69-76

Fedrick lA, Cromie SJ, Dulley JR (1986) The effects ofincreased starter populations, added neutral pro-teinase and elevated temperature storage on Ched-dar cheese manufacture and maturation. NZ J DairySei Teehnol21, 191-203

Folkertsma B, Fox PF, McSweeney PLH (1996) Accel-erated ripening of Cheddar cheese at elevated tem-peratures. Int Dairy J 6, 1117-1134

Fox PF (1963) Potentiometric determination of salt incheese. J Dairy Sei 46, 744--745

IDF (1964) Determination of the protein content ofprocessed chee se products (Standard 25). Interna-tional Dairy Federation. Brussels

IDF (1982) Chee se and processed cheese - Determi-nation of total solids content (Standard 4a). Inter-national Dairy Federation. Brussels

URS (1955) Determination of the percentage of fat incheese (Irish Standard 69). Institute for IndustrialResearch and Standards. Dublin

Jordan KN, Cogan TM (1993) Identification andgrowth of non starter lactic acid bacteria in IrishCheddar cheese. Irish JAgrie Food Res 32, 47-55

Kanawjia SK, Nageswara Rao K, Singh S, Sabikhi L(1993) Role of lactobacilli in chee se - A review.Indian J Dairy Sei 46, 187-197

Khalid NM, Marth EH (1990) Lactobacilli - Theirenzymes and role in ripening and spoilage ofcheese: A review. J Dairy Sci 73, 2669-2684

Kleter G (1977) The ripening of Gouda cheese madeunder strictly aseptic conditions. 2. The comparisonof the activity of different starters and the inllu-

573

ence of certain Laetobaeillus strains. Neth MilkDairyJ31,177-187

Law BA, Hosking ZD, Chapman HR (1979) The effectof sorne manufacturing conditions on the develop-ment of Ilavour in Cheddar cheese. J Soc DairyTechnoI32,87-90

Martley FG, Crow VL (1993) Interactions betweennon-starter microorganisms during cheese manu-facture and ripening. Int Dairy J 3, 461--483

Martley FG, Lawrence RC (1972) Cheddar cheeseIlavour. II. Characteristics of single strain startersassociated with good or poor Ilavour development.NZ J Dairy Sei Technol7, 38--44

McSweeney PLH, Fox PF, Lucey JA, Jordan, KN,Cogan TM (1993) Contribution of the indigenousmicrollora to the maturation of Cheddar chee se.Int Dairy J 3,613--634

Morgan S, O'Donovan C, Ross RP, Hill C, Fox PF(1995) Significance of autolysis and bacteriocin-induced Iysis of starter cultures in Cheddar cheeseripening. In: Proc. 4th Cheese Symposium (CoganTM, Fox PF, Ross RP, eds) Teagasc, Fermoy Co"Cork, 51--60

O'Connor CB (1974) The quality and composition ofCheddar cheese. Effect of various rates of salt addi-tion. III. Irish Agric Creamery Rev 27, 11-13

Peterson SD. Marshall RT (1990) Nonstarter lacto-bacilli in Cheddar cheese: A review. J Dairy Sei73, 1395-1410

Rogosa M, Mitchell JA, Wiseman RF (1951) A selec-tive medium for the enumeration of oral and faecallactobacilli. J Baeterio/62, 132-133

Terzaghi BE. Sandine WE (1975) Improved media forlactic streptococci and their bacteriophages. App/Mierobio/29,807-813

Thomas ID (1987) Cannibalism among bacteria foundin cheese. NZ J Dairy Sei Techno/22, 215-219

Thomas TD, Pearce KN (1982) Inlluence of salt onlactose fermentation and proteolysis in Cheddarcheese. NZ J Dairy Sei Teehno/ 16, 253-259

Turner KW, Thomas TD (1980) Lactose fermentationin Cheddar cheese and the effect of salt. NZ J DairySei Teehnol 15,265-276

Visser FMW (1977) Contribution of enzymes fromrennet, starter bacteria and milk to proteolysis andIlavour development in Gouda cheese. 1. Descrip-tion of cheese and aseptic cheesemaking techniques.NethMilkDairyJ3I,12D-133